What does it mean to look at an algorithmic system as a game?

One of the games designed at the
2nd NOS-HS workshop for Nordic
Perspectives on Algorithmic
Systems. Photo copyright:
Salla-Maaria Laaksonen

It is not uncommon to hear the phrase “gaming the system” when someone fools an algorithm, but game as a metaphor can go beyond this perspective. Besides calculation, games are also about meanings and design.

We recently published an article called “Gaming Algorithmic Hate-Speech Detection: Stakes, Parties, and Moves” with Salla-Maaria Laaksonen and Airi Lampinen in the Social Media+ Society journal. In it, building on Erving Goffman’s work, we use the game metaphor to study the implementation of and subsequent resistance to an automatic hate-speech detection system. This blogpost describes my perspective on how games can be used as a metaphor to approach algorithmic systems from different angles. It is not meant as a comprehensive list, but more as an example of some of my thoughts on the topic during my Ph.D. journey so far.

I leave the concept of an algorithmic system intentionally vague here. You may consider recommender systems that suggest content to you in Netflix or Youtube as one type of an example, and social network services such as Facebook as another one. Generally, many of the information systems we interact with on a daily basis can be thought of as algorithmic systems. Thus, a tongue in cheek way of describing algorithmic systems in the context of this blog post could be “any information system that uses algorithms in a way that interests the author of this blog post”.

A game as a general metaphor

This is a category where I would group approaches such as the one which we took in the aforementioned article: if everyday life is (sometimes) a game, what can we learn if we approach it analytically as such? At least two different ways of understanding this type of approach can be identified: ideas about gaming the system, which often refer to behavior where someone is seen to “cheat” an algorithm and, on the other hand, following the rules, where individuals maximize their gains by acting in line with how they consider the designers of the system(s) wish them to act. Cotter’s article “Playing the visibility game: How digital influencers and algorithms negotiate influence on Instagram” wonderfully illuminates and also problematizes this distinction when discussing simulated and relational influence in terms of how Instagram influencers attempt to increase their visibility. Additionally, in her recent article “Algorithmic Experts: Selling Algorithmic Lore on YouTube”, Bishop notes that algorithms may be treated as ‘games’ by those attempting to figure out how they could be used to one’s advantage.

This line of consideration also raises questions about who or what are players in a particular game playing for. Companies specializing in search engine optimization are obviously trying to make their clients’ pages generate more traffic, and fans of musical groups might be trying to give the target of their fandom more visibility. We certainly do not play only for ourselves.

Algorithmic systems as world creating

For me, this second way of looking at algorithmic systems – algorithmic systems as world creating – entails looking at what kind of a ‘micro-cosmos’ of meanings the encounter with the system holds. In a game of chess, pieces have different values for the player based on their shape, but the relationship between the shape and the value only makes sense in the context of chess. Goffman points out that other encounters share this element of encounter-specific meanings: “It is only around a small table that one can show coolness in poker or the capacity to be bluffed out of a pair of aces; but, similarly, it is only on a road that the roles of motorist and pedestrian takes on full meaning”. Encounters, then, are world-creating events, and encounters with algorithmic systems are no exception.

Algorithmic systems have different kinds of transformation rules. The concept originates from Goffman, but I find Di Filippo’s use of it in his book chapter “MMORPG as Locally Realized Worlds of Action” easier to grasp than the original definition. Di Filippo states that these transformation rules refer to “the fact that individuals adapt resources to match the relevance of the situation”. In the aforementioned chapter, Di Filippo uses the concept to analyze how the world created in fantasy books is transformed to serve as a backdrop for a video game. When considering recommender systems, transformations occur on how behavior should be understood: clicks or other forms of behavior such as decisions to buy something are transformed into recommendations. If a couple of strangers in front of us in a cafĂ© buy the same kind of coffees, we most likely do not consider it as a recommendation: however, it could very well be transformed into one based on the data collected from the transaction that happens between the clients and the cafĂ©. We have followed this line of inquiry with Airi Lampinen in a study that drew from Goffman’s Frame Analysis by analyzing interviews of users and the head designer of a recommender system that used reading time to generate its recommendations.

Game-likeness from a design perspective

This third category – game-likeness in the context of design – focuses on the design perspective of algorithmic systems, or more specifically, what algorithmic systems that are not intended as games may share with the design of games. Algorithmic systems may incorporate elements that make them “game-like” or gamified. Instagram and Facebook quantify “likes” other users can give to one’s content, making these systems potentially more enthralling for their users. From a perhaps more serious perspective, Chan (2019) has pointed out in his article “The rating game: The discipline of Uber’s user-generated ratings” that the on-demand taxi service Uber’s customer reviews make drivers attempt to maximize positive ratings as their livelihoods may be on the line: get enough negative reviews and you won’t be getting customers anymore. 

Game design can also be a method that ties together some of the elements from the first two categories. In the 2nd NOS-HS workshop for Nordic Perspectives on Algorithmic Systems, Michael Hockenhull and Mace Ojala organized a session where the participants designed tabletop-games from empirical research cases Bastian JĂžrgensen, CĂŠcilie Laursen, Silja Vase and Rikke Torenholt were working on and were kind enough to let us use as starting points for the games. This activity was inspired by Dumit’s article “Game Design as STS Research”. The process of designing a game based on an algorithmic system forced one to consider both the calculative nature of interactions individuals may have with such systems and the ways these interactions could be transformed into a playable format.

Conclusion

As pointed above, games can be used as a metaphor to illuminate different kinds of things about algorithmic systems. The focus can be placed on the strategic nature of everyday dealings with them, the set of meanings interactions with them contain, or the design aspects that may mimic those we encounter in actual games. One could probably discover further perspectives that the concept of a game might afford, but these are the three that I have identified in the extant literature and found productive for my own research.

Jesse Haapoja is a Ph.D. student in Social Psychology at the University of Helsinki who has the privilege of working on topics such as the one presented here in the Kone Foundation funded project “Algorithmic Systems, Power, and Interaction”.

Thanks to Airi Lampinen for comments on a draft of this blog post

HybridejĂ€ mainenarratiiveja, tunteella ja teknologialla – vĂ€itöstilaisuus 16.6.2017

Screen Shot 2017-06-06 at 00.19.27VTM Salla-Maaria Laaksonen eli allekirjoittanut vĂ€ittelee 16.6.2017 kello 12 Helsingin yliopiston valtiotieteellisessĂ€ tiedekunnassa aiheesta “Hybrid narratives – Organizational Reputation in the Hybrid Media System“. Tervetuloa mukaan vĂ€itöstilaisuuteen kuulemaan akateemista debattia organisaatiomaineesta ja verkkojulkisuudesta! Alla lyhyt yhteenveto tutkimuksesta.

***

Tutkin vÀitöskirjassani sitÀ, miten yrityksiÀ ja muita organisaatioita koskevat mainetarinat muodostuvat hybridissÀ mediatilassa. Tutkimusongelma on kaksitahoinen: tutkin, miten uusi viestintÀympÀristö vaikuttaa organisaatiomaineen muodostumiseen, ja toisaalta sitÀ, minkÀlaisia kognitiivisia ja emotionaalisia vaikutuksia maineella ja mainetarinoilla on. Hybridi mediatila on viestinnÀn tutkimuksen tuoreehko kÀsite (Chadwick 2013), joka pyrkii ymmÀrtÀmÀÀn nykyistÀ mediamaisemaa. Hybridiys viittaa eri mediamuotojen sekoittumiseen: sosiaalisen median ja perinteisen median sisÀllöt ja muodot elÀvÀt verkkojulkisuudessa vahvasti sekoittuneena.

Tarkastelen vĂ€itöskirjassani hybridia mediatilaa tarinankerronnan paikkana. TĂ€stĂ€ nĂ€kökulmasta jokainen blogikirjoitus tai twiitti on pieni kertomus, jollaisia teknologia kutsuu meitĂ€ kertomaan arjesta ja kokemuksistamme. Monet kertomuksista kĂ€sittelevĂ€t suorasti tai epĂ€suorasti yrityksiĂ€ ja muita organisaatioita – jolloin ne ovat mÀÀritelmĂ€llisesti mainetarinoita. NiitĂ€ on jaettu arjessa aikaisemminkin, mutta teknologia mahdolistaa uudenlaista tarinankerrontaa: tarinat leviĂ€vĂ€t lĂ€hipiiriĂ€ laajemmalle, ne arkistoituvat, ja nistĂ€ tulee etsittĂ€viĂ€ ja muokattavia.

Osa verkon teknologioista toimiikin tarinankerronnan apuvÀlineinÀ hyvin erityisellÀ tavalla: ne jÀrjestÀvÀt, kuratoivat ja muovaavat kertomuksia yhdistÀmÀllÀ erilaisia tarinanpalasia yhteen nÀkymÀÀn. NÀin toimii esimerkiksi joukkovoimin yllÀpidetty tietosanakirja Wikipedia tai verkon sisÀltöÀ penkovat hakukoneet. SIksi hybridissÀ mediassa maineen tarinankertojina toimivat sekÀ ihmistoimijat ettÀ teknologia yhdessÀ. VÀitöskirjani pohjalta esitÀnkin, ettÀ teknologia muuttaa niitÀ tapoja, joilla sidosryhmÀt kertovat tarinoita organisaatioista. Verkkojulkisuuden alustoilla syntyvissÀ tarinoissa sekoittuvat paitsi eri mediamuodot, myös faktat ja mielipiteet sekÀ rationaalinen ja emotionaalinen sisÀltö.

VÀitöskirjani korostaakin tunteiden merkitystÀ maineelle. Niin maineen tutkimuksessa kuin erilaisissa mainemittareissakin on perinteisesti keskitty rationaalisiin ominaisuuksiin: tuotteiden laatuun, johtajuuteen, taloudelliseen menestykseen. Maine nÀyttÀisi kuitenkin olevan yhtÀ paljon myös emotionaalinen kÀsite. Organisaatioita kÀsittelevÀt kertomukset verkossa ovat hyvin tunnepitoisia: yritysten kanssa ihastutaan ja vihastutaan, niiden ympÀrille rakentuu faniyhteisöjÀ ja vihaisia kohuyhteisöjÀ. Teknologian ominaisuudet emoji-hymiöistÀ tykkÀÀ-nappulaan myös kannustavat ilmaisemaan tunteita.

EikÀ tunteissa ole kyse vain ilmaisusta. VÀitöskirjan osatutkimuksessa osoitettiin, ettÀ hyvÀ ja huono maine nÀkyvÀt eri tavoin koehenkilöiden kehollisissa reaktioissa, kun he lukevat yritystÀ koskevia verkkouutisia tai verkkokommentteja. Maine on siis myös tulkintakehys: tiedostamaton, kehollinen reaktio, joka ohjaa ihmisen toimintaa esimerkiksi ostoksilla valintatilanteessa.

Mainetutkimuksen nÀkökulmasta rakennankin työssÀni uudenlaista kulmaa mainetutkimukseen. Mainetta on perintisesti tutkittu joko organisaation taloudellisena voimavarana tai tulkinnallisena elementtinÀ sidosryhmien mielissÀ. TÀssÀ työssÀ mÀÀrittelen maineen viestinnÀllisenÀ ilmiönÀ, joka on olemassa yksilöiden tulkintakehyksenÀ sekÀ sosiaalisesti rakentuneina narratiiveina. Mainenarratiiveilla on kuitenkin myös mitattavia vaikutuksia niitÀ lukeviin ihmisiin ja heidÀn tulkintakehyksiinsÀ. Siksi sekÀ maine ettÀ mainetarinat ovat organisaatioille aineetonta pÀÀomaa.

VÀitöskirja koostuu viidestÀ artikkelista ja yhteenvetoluvusta. Artikkeleissa on kÀytetty neljÀÀ eri aineistoa: viestinnÀn ammattilaisten haastatteluja, sosiaalisen median verkkokeskusteluaineistoja, Wikipedia-aineistoa sekÀ psykofysiologisia mittauksia. NÀin ollen tutkimus yhdistÀÀ metodisesti laadullista, narratiivista analyysia kokeelliseen tutkimukseen.

***

TLDR; “Hybridi mainetarina syntyy kun đŸ˜© ja đŸ‘Ÿ yhdessĂ€ kĂ€yttĂ€en apunaanđŸ“±đŸ’», muodostavat 📜💌📜 , jotka verkkojulkisuudessa đŸ’Ÿ ja 📱 ja joilla on 📉 vaikutuksia 🏭🏹 🏱:lle.” (ref. Your Research, emojified)

VÀitöskirjan elektroninen versio on luettavissa E-thesis -palvelussa.

VÀitöskirjaa ovat rahoittaneet Liikesivistysrahasto ja Tekes.

Suomi24 Data Science Hackathon – results and afterthoughts

The availability of large data sets and digital material is changing the landscape of research within social sciences and humanities. At the same time, tools and the understanding necessary to utilize such data are often lacking. To tackle this problem, during the last weekend of May we organized a Data Science hackathon around a newly opened data set of Suomi24, the largest online discussion forum in Finland with 1.9 million monthly visitors.

The hackathon was organized by the Citizen Mindscapes research collective, University of Helsinki, Futurice Oy and Aller Media Oy. The event was also part of Nordic Open Data Week and organized in cooperation with Open Knowledge Foundation. The main goal of the event was to allow researchers and coders work together and find new ways of collaborating in the field of data science. We built four different teams consisting of coders and researchers to figure out research problems and create solutions and demos to find their answers.

The dataset used in the event was the almost entire database of Suomi24 online forum discussions ranging from 2001 to 2015, consisting of hundreds of thousands of posts and altogether over 123 million words – a set of data rather impossible to study comprehensively using traditional methods from social science or humanities. Below is a summary of the work and results discovered by the teams.

Rhythms of Human Life in Suomi24

This team was interested in the life cycle of topics in Suomi24. A typical way of studying topics is creating a list of words and querying the data with the words. As one exercise this team tracked the conversations related to jealousy using a list of fifteen related words. They noted that in general the talk about jealousy has increased during the time span of the data. Maybe people were not so used to talk about personal issues online but year by year it is getting more common? Further, the analysis shows that jealousy words peak during January and in May; on the contrary in December discussions on the topic are rare. The team hypothesized that this relates to the well-known phenomenon of finding a summer fling, or the aftermath of all the Christmas parties.

b0hp85WFY244R2vE3HiXcCgd1qz8rSyCIypof7WUGGQ

User Modeling and Micro Level Interactions

This team focused on tracking down different interaction types, recognizing positive/negative discussions, and finding out what words or linguistic features are predicting longer discussion threads. In essence these questions directly relate to a very practical problem of how to create interaction in the online sphere and produce text so that the writer can create engagement. The team decided to simply measure this using the length of the thread as the dependent variable, and using MDL (Minimum Description Length) started searching for the linguistic features that are typical to long or short conversation threads. Limiting the analysis to conversation sections related to babies and society, they identified some discreet words, topics and features of the text that are typical for short and long threads (see table below).

 SHORT LONG
baby section: inconvenient topics (pregnancy, test, symptoms, miscarriage, periods)
society section: god, work, human
baby section: boy, kid, man, girl, mother, movie
society section: Jesus, forest, baptize
asking, short sentences, question mark, words indicating uncertainty (mikÀ mutta vai jos), colloquialism subordinate clauses, certain conjunctions (ettÀ, vaikka, ja), quotations, commas

Forecasting the Economy

Our forecasting team decided to study what words and topics get accentuated during a financial downturn, and to check whether the online discussions could be used as a tool to predict the economical situation. The theoretical idea behind this question comes from John Maynard Keynes’s notion of animal spirits; the instincts, fears and emotions ostensibly influence and guide human behavior, and through that also affect the economic cycle. In order to answer their questions the team obtained additional data sets regarding Finnish GDP and private household consumption from the National Statistics Finland. An index to measure economic uncertainty in the discussions by a set of key words was created using previous studies as a source. An OLS regression model was tested but didn’t have large explanatory power with this data set. Nevertheless, in the next part of the analysis the team  identified the words whose frequencies rose during the months of the crisis years 2008 and 2009. So, if the economical situation is going down, what are the words people use more often? The identified words were: bar, mother-in-law, poem, weapon, bank, electricity, unemployed, lonely, Easter, girlfriend. We do hope these words are not related to a single story!

Cats versus Dogs

Our last team decided to solve the old Internet dilemma of cats versus dogs once and for all. It is well known that Internet belongs to cats. But how about Suomi24? Are cats also the most prominent animals there? Different statistics were extracted from the data, but the situation kept looking bad for cats: dogs are mentioned more often across the data. Also the amount of users who talk about dogs versus cats is larger. A final analysis was conducted to see whether other topics that cat/dog persons talk about actually differ. The results show what cat people do talk more about mathematics, where as dog persons talk about poop. This whole exercise of course was just a humorous example of what to do with the data, and how to twist the data so that a needed answer can be found – it is just a matter of what to measure. A critical point to note is thus that one should be cautious of different black boxes of data analytics: there might have been other statistics behind the ones that you are shown.

Screen Shot 2015-06-03 at 16.02.51

Some afterthoughts

Apart from the fantastic results from the demos the whole event of course was a learning experience. Most important observation is the need for multidisciplinary knowledge and skills within the teams. Without a more general, wider knowledge about the societal phenomena that are affecting the creation of such social big data in the first place it is not possible to draw relevant conclusions. Our hypotheses of the jealousy discussion, for instance, are pure speculations for now, but probably a dwell into social psychology research on the topics would take us lot further.

Also there’s a clear need to better understand the context of the words studied, as their meaning can be heavily dependent on that. Based on the cat vs. dogs analysis, for instance, we can’t say whether the discussions about cats or dogs are actually pro-cats or pro-dogs or are people actually just complaining about the neighbors pet – this would need deeper analysis regarding the context and tone of the messages.

And of course during two days you probably will not learn that many new skills but rather utilize the old ones in a new context. So no two-day magic crash courses to python coding actually happened, but hopefully some broadening of mindscapes for researches both in social and computational sciences!

  • The Suomi24 data set can be explored through FinCLARIN’s Kielipankki Korp-interface. Full data set is available for download for research purposes.
  • Follow Citizen Mindscapes researcher collective in Twitter.
  • Team members: Rhythms team Pasi Karhu, Limae Phuah, Omar El-Bagawy, Jaakko Suominen, Krista Lagus, Minna Ruckenstein; User Modeling team Antti Rauhala, Krista Lagus; Forecasting team Kimmo Nevanlinna, Timo NikkilĂ€, Joonas Tuhkuri; Cat vs. Dogs group Matti Nelimarkka, Salla-Maaria Laaksonen.

This post is a cross-posting from Opennorcids.org

Ajatuksenvirtaa Helsinki Digital Humanities DaystĂ€

Joulukuun 3. vietettiin Helsinkin yliopistokollegiumin jÀrjestÀmÀÀ Digital Humanities -pÀivÀÀ. PÀivÀ oli tarkoitettu tutkijoille ja muille aiheesta ja uudesta tutkimualueesta kiinnostuneille. VÀkeÀ löytyikin salin tÀydeltÀ! Ehdin itse olla paikalla vain aamupÀivÀn, mutta tÀssÀ muutamia tuntoja.

PĂ€ivĂ€n aluksi Arto Mustajoki lĂ€hti etsimÀÀn “digitaalisen humanismin” mÀÀritelmÀÀ. Lainausmerkit erityisesti siksi, ettĂ€ toimivaa suomennosta termille ei oikein ole löytynyt (ks. mielenkiintoista keskustelua aiheesta Qaiku-ryhmĂ€ssĂ€). Siksi tĂ€ssĂ€ blogikirjoituksessa taidan pysytellĂ€ lyhenteessĂ€ DH.

Wikipedia mÀÀrittelee DH:n seuraavasti:

Digital humanities is an area of research and teaching at the intersection of computing and the disciplines of the humanities. Developing from the fields of humanities computing, humanistic computing, and digital humanities praxis digital humanities embraces a variety of topics, from curating online collections to data mining large cultural data sets. (3.12.2014)

LÀhtökohta pÀivÀlle oli vahvasti humanistisissa tieteissÀ. Aamulla esiteltiin esimerkkejÀ muun muassa Shakespearen teosten analyysista kieleen keskittyvÀllÀ lattice analysis -menetelmÀllÀ (lattice on suomeksi hila, mutta en ole nÀhnyt kÀytössÀ analyysimenetelmÀlle varsinaista suomennosta?) japanilaisen taiteen hakemistoihin ja historiallisiin karttoihin ja karttatietokannan rakentamiseen osin joukkoistettuna.

Itselleni avartavin oli paraikaa Yliopistokollegiumissa vierailevan Caroline Bennettin (Sussex University) puheenvuoro, jossa kÀsiteltiin paitsi lyhyesti Sussexissa ensi vuonna starttaavaa Humanities Labia, myös yleisesti digitalisoitumisen vaikutusta tieteeseen ja kÀsityskykyymme. Bennettin mukaan digitalisoitumisessa tai DH:ssa on kysymys tutkimusmateriaalin muutoksesta, mutta samalla muutos vaikuttaa siihen, miten ylipÀÀnsÀ nÀemme ja koemme tutkimuskohteemme. HÀn muistutti, ettÀ kriittinen asenne olisi syytÀ sÀilyttÀÀ, mutta lÀhteÀ rohkeasti kokeilemaan monitieteisesti uusia menetelmiÀ avoimesti ja verkostuen. Mutta peruspohja on tieteessÀ: valmiit työkalut ovat aina vain työkaluja, jotka antavat vastauksia, mutta eivÀt kerro meille kysymyksiÀ.

Kytkeytyy siis hyvin big data -pöhinÀn ympÀrillÀ kÀytÀvÀÀn tutkimuksen perusongelmaan: dataa on, ja siitÀ voidaan kaivaa vaikka mitÀ, mutta so what? Kuka esittÀÀ oikeat kysymykset ja mitÀ merkitystÀ niillÀ todella on? Kuten vierustoverini osuvasti luennolla kysÀisi: jos Shakespearen tuotannosta ei oltu ilman data-analyysia osattu löytÀÀ kolmea muusta tuotannosta selkeÀsti erottuvaa teosta kaikkien luettujan vuosisatojen ja tutkimusenkaan jÀlkeen, onko eroilla oikeasti jotain vÀliÀ?

Juuri siksi monitieteisyys ja datasokeuden vÀlttÀminen ovat niin tÀrkeitÀ asioita. Itse nÀkisin, myös jonkun luennoiman esittÀmÀn ajatuksen mukaan, ettÀ digitaaliset automaattiset menetelmÀt ovat hyvÀ keino exploratiivisesti sukeltaa dataan, mutta sen jÀlkeen olisi hyvÀ paneutua yksityiskohtiin laadullisesti tai mÀÀrÀllisesti; esimerkiksi selvittÀÀ tÀsmÀllisemmin, miten ne Shakespearen kolme teosta ovat erilaisia.

Iloinen uutinen on joka tapauksessa se, ettĂ€ Digital Humanities on yksi ehdotettu painopisteala humanistiselle tiedekunnalle ja heillĂ€ on aikomus perustaa DH Lab myös Helsingin yliopistoon, aluksi todennĂ€köisesti kevyellĂ€ organisoitumisella ja monitieteisesti. Toivottavasti kuulemme tĂ€stĂ€ pian lisÀÀ! (Mustajoelta terveisiĂ€, ettĂ€ potentiaaliset yliopiston ulkopuoliset rahoittajat ovat kuulemma tervetulleita – yliopistorahoituksen ehdoilla kun mennÀÀn 🙂

Edit 4.12.2014: LisÀtty linkit Qaiku-keskusteluun ja Storify-koosteeseen sekÀ virke toiseksi viimeiseen tekstikappaleeseen.